Create a vector function from a scalar function in R

Question

Here is my function. It takes certain values on [-5,-3] and [3,5] and 0 elsewhere. The function is symmetric around the origin.

f<-function(x){
    y=0 
    if (x>=-5 && x<=-3){
        y=3*(1-(x+4)^2)/8
    }
    if (x>=3 && x<=5){
        y=3*(1-(x-4)^2)/8
    }
    return(y)
}

Okay so I used the function sapply (bins, f)

where bins = seq(-5,5,by=0.05). This worked fine!

but when I tried to do f(bins), I got this ridiculous answer. It was correct for the [-5,-3] range.

My guess for this is that when the function f checked the first if condition, it checked it only for the first value of the bins vector, so for (-3,5] range, it incorrectly used the formula only intended for [-5,-3].

I am trying to get a way to draw a curve for these points, but when I used curve function, the curve is drawn using the wrong values we would get by using f(bins)

Can someone please tell me how to fix this?

Answer 1

You would have to Vectorize the function f to get it to apply over a vector:

f = Vectorize(f)
print(f(bins))

Note that you could have also just used curve with sapply:

curve(sapply(x, f), from=-5, to=5)

Finally, if you wrote the function with ifelse like so:

f = function(x) {
    ifelse(x >= -5 & x <= -3, 3*(1-(x+4)^2)/8, ifelse(x>=3 & x<=5, 3*(1-(x-4)^2)/8, 0))
}

That would allow it to work on vectors without needing Vectorize.